proteus anguinusparkelj (urodela: proteidae) project/a black...

A black, non-troglomorphic amphibian from the karst of Slovenia:

Proteus anguinus parkelj n. ssp. (Urodela: Proteidae)

B. Sket & J.W. Arntzen

1Department of Biology, Biotechnical Faculty, University of Ljubljana, Vecna pot 111,

P.O. Box 486/492, 61000 Ljubljana, Slovenia; 2Binnen Oranjestraat 10, 1013 JA Amsterdam,

The Netherlands

Keywords: Urodela, Proteus, taxonomy, morphology, allozyme analysis, ecology, distribution

Abstract

A morphologically distinct cavernicolous salamander Proteus

anguinus from southeastern Slovenia (Bela Krajina) is described

as P. a. parkelj ssp. n. It differs from P. a. anguinus in a dark

pigmentation, fully developed eyes, a skull with broader and

shorter bones and fewer teeth, a voluminous jaw musculature

that gives the head a bulky appearance, a proportionallylonger

trunk with a higher number of vertebrae, shorter extremities,

and a shorter tail. Most of these traits are considered to be

plesiomorphic character states. An allozyme analysis over 40

loci has shown the new dark pigmented taxon to be genetically

similar to a white and troglomorphic neighbouringpopulation

from Stična (DNei= 0.23). Both populations in turn are geneti-

cally dissimilar to a geographically more distant population

from Postojna (D Nei= 0.49). The observed level of genetic

differentiation suggests that western and southeastern Slovenian

populationsform separate lineages since the uppermost Miocene

but conservatively hitherto only a single species is recognised.

Thenew taxon is onlyknown from asmall area and may be rare.

P. a.parkelj now under strict legal protection, is threatened by

industrial pollution.

Résumé

Un Protée remarquablede Slovénie du sud-est (Bela Krajina) est

décrit comme Proteus anguinusparkelj ssp. n. Celui-ci se distin-

gue de la sous-espèce nominative par la pigmentation très fon-

cée, les yeux bien développés, le crâne à os plus courts et élargis,

un nombre inférieur de dents, la musculature des mandibules

plus fortement développée (ce qui donne à la tête un aspect plus

massif), le corps relativement plus allongé et à nombre supérieur

de vertèbres, ainsi que par les pattes et la queue plus courtes. La

plupart de ces caractères sont considérés comme étant plésio-

morphes. L’analyse des allozymes à 40 loci montre que ce taxon

à pigmentation foncée ressemble au point de vue génétique

(DNei

= 0.23) à la population troglomorphedépigmentéed’une

localité située à proximité (Stična). Cependant, ces deux popula-

tions diffèrent génétiquement d’une population géographique-

ment plus distante (Postojna; DNei

= 0.49). Le niveau de

différenciation génétique montre que les populations occiden-

tales et sud-orientales représentent des lignées distinctes depuis

le Miocène supérieur; on considère tout de même qu’elles appar-

tiennent à une même espèce. Le taxon nouvellement décrit est

connu d’un territoire fort restreint, et il est apparemment rare;

bien que protégé par la loi, P. a. parkelj n. ssp. est menacé par

la pollution industrielle.

1 History, habitat, and distribution

The known distribution of Proteus anguinus en-

compasses the Dinaric karst almost entirely, from

the lower reaches of the Isonzo-Soca River in Italy

in the northwest to the river Trebiänjica in Hercego-

vina in the southeast. Altogether almost 200 locali-

ties are known (Sket, 1983; Sket & Aljanëiü, in

prep.).

Proteus is highly specialised troglomorphic,

Bijdragen tot de Dierkunde, 64 (1) 33-53 (1994)

SPB Academie Publishing bv, The Hague

The history of the name Proteus goes back to Val-

vasor (1689) who mentioned the species in his trea-

tise "Die Ehre des Herzogtums Crain". With its

discovery in the Postojna caves in Central Slovenia

by Jersinovic von Loewengreif in 1797 (Fitzinger,

1850) Proteus became the first known specialised

cave animal.

The normal habitatof Proteus is cave waters, but

most places where it was found were in the open

when occasionally specimens were washed out of

the caves after heavy rainfall. J.N. Laurenti, who

formally described the species in a monotypic genus

in 1768, obtained his material from such a surface

location in Cerknica near the Postojna caves.

B. Sket & J. W. Arntzen - A black Proteus34

strictly neotenic and shows limited morphological

variation. Normally Proteus is blind and at the

most weakly pigmented. However, it is able to syn-

thesise dark pigments and reared in the light it may

gradually become dark or even black (Aljancic et

al., 1986). During larval development, the initial

eye development is retarded; lateron in life the eyes

degenerate and become covered by skin (Durand,

1973). Attempts to induce metamorphosis failed

(Pehani & Seliskar, 1941).

No obvious characters exist that suggest strong

intraspecific differentiation. The only serious at-

tempt to elaborate upon the taxonomy of Proteus

was by Fitzinger (1850) who described seven spe-

cies under the genus name Hypochthon (see also

Schreiber, 1912). Of all those taxa only H. zoisii has

been accepted by some (Mertens & Müller, 1940) at

the subspecies level as Proteus anginus zoisii. Pro-

teus anguinus; is the only living representative of the

genus. The most likely sister taxon to Proteus is the

genusNecturus from the New World (Duellman &

Trueb, 1986).

The first documented finding of a "black" Pro-

teus in nature is that by membersof the Institute for

Karst Research (Slovenian Academy of Sciences

and Arts). One black specimen with developed eyes

and unusual body proportions was caught in the

Doblicica spring, near Crnomelj in southeastern

Slovenia on October 18, 1986. As Proteus is pop-

ular among Slovenes, this finding resulted in a lot

of publicity, mostly at the popular level (AljanCiC

et al., 1986; Istenic & Bulog, 1986; IsteniC, 1987;

Hodalic, 1993). Another black specimen was sight-

ed and photographed by Miss M. Zlokolica be-

low the Jelsevnik springs, some 2.5 km north of

Doblicica, on April 21, 1990. In the period up to

1992 a sufficient numberof specimens (n = 15) was

obtained to allow morphometric and other investi-

gations.

For the purpose of the present paper we will re-

fer to densely pigmented Proteus as "black" and

to unpigmented and slightly pigmented ones as

"white".

2 Material and methods

used or mentioned in this study originate

mainly from the following Slovenian localities: (1) the NaTrati

Samples of Proteus

Fig. I. Southern Slovenia with the area of Dinaric Karst indicated by a denticulated line. Solid squares refer to documented records

of white P. a. anguinuspopulations; solid diamonds refer to documented records of P. a. parkelj; open symbols refer to reports that

require confirmation. The insert highlights the Bela Krajina, with black dots documentingplaces of access to hypogean waters such

as springs and caves.Scale bar is 5 km. D, Dobličica; J, Jelševnik; P, Planina (Postojna and Planinska Jama); S, Stična, with the springs

Vir and Rupnica. Three vertical bars indicate (from left to right) the cities Trieste, Ljubljana, and Črnomelj.

35Bijdragen tot de Dierkunde, 64 (1) - 1994

spring at Jelsevnik nearCrnomelj in the Bela Krajina (inthe text

usually referred to as "JelSevnik"), (2) the Vir and Rupnica

springs, both at Sticna, 1.7 km apart, 25 km E.S.E. of Ljubljana

(togetherreferred to as "Sticna"), (3) the Planinska Jama cave

at Planina, near Postojna which forms part of the Postojna-

Planina cave system (referred to as "Planina") (Fig. 1).

Only specimens that were damaged or that died in captivity

have been preserved, except for specimens that were sacrificed

for biochemical taxonomie analysis (see below). As a conse-

quence the quality of most of the material is rather poor. For

histologicalstudies the specimens were fixed in a 2% solution of

glutaraldehydeand later transferred to 70% ethanol. One skele-

ton was prepared by maceration of the carcass in aweak solution

of KOH and its skeleton then dyed by alizarin. Other specimens

or parts thereof have been cleared in KOH and glycerol, stained

with alizarin and preserved in glycerol with thymol, following

Humason (1967). Eight black Proteus are preserved in 70%

ethanol and at present six are kept alive at the Department of

Biology of the Biotechnical Faculty, University of Ljubljana.

Some osteological details were studied on cleared specimens

from Planina and from Luknja near Novo Mesto. In addition,

tail tips of some specimens from Vir, Luknja, and Planina were

cleared to allow an osteological study. Data on pigmentation

were taken from live as well as from preserved specimens. Eyes

were measured in some intact specimens and in a skinned head.

Data onthe structure of the eyes and skin were obtained by elec-

tron microscopy (Bulog, 1992).The morphology of the gills was

studied in detail but since preliminary observations pointedto a

marked phenotypic plasticity, the quantitativeaspects were not

further taken into consideration. X-ray photographswere taken

by standard clinical equipment. For most specimens settings

were 31-32 kV at 120 mA on a Siemens "Selenos-4", with a

glassplate below servingas a filter. In newly collected specimens

listed in Table II, photographs were taken using a Siemens

"Polydoros 80" apparatus set at 40-41.5 kV and 1.6—2.5 mA.

We did not succeed in obtaining useful pictures of the tiny tail

vertebrae, neither by X-ray photography, nor by nuclear mag-

netic resonance.

A specimen of Necturus sp. of unknown provenance and its

X-ray image were used for morphological comparison of the

eyes and the backbone. Two adult specimens of Triturus alpes-

tris (Laurenti, 1768) from the Kamniske Alps in Slovenia and

two specimens of neotenic phenotype from Trnovicko Jezero,

Maglic Mts., Montenegro were cleared for the same purpose.

A series of 82 individuals, includingspecimens stored in the col-

lections ofthe Naturhistorisches Museum, Vienna, and the Uni-

versity of Ljubljana, originatingfrom four localities (Jelsevnik,

Planina, Rupnica, and Vir) were subjected to the measurement

of the following eight variables: total length from the tip of the

snout tothe tail tip (L); head length from the tip of the snout to

the line connecting the bases of the most anterior gills (Lc); head

width at the point where head width is largest (Ltc); pre-pedal

length from the tip of the snouth to the anterior legs (LaP);

trunk length between the hind side of the anterior and the fore

side of the posterior leg (LiE); the length ofthe anterior and the

posterior legs from their point of insertion to the tip of the

longest toe onthe perpendicularlystretched leg (PaL, PpL), and

finally tail length from the hind side of a posterior leg to the tip

of the tail (Led). Measurements were taken to the tenth ofa mm

using vernier callipers. Presented values are rounded off totheir

integer values. It should be noted that in soft tissues and in con-

tracted preserved material the precision of the measurements

may be diminished.

Two meristic measurements were also taken: (1) In fresh and

alcohol-preserved material the number of myomeres (Myo) was

counted between both pairs of legs. (2) The number of vertebrae

(Ver) from the atlas up to and including the last vertebra before

the pelvic girdle(ilium) was determined from X-ray photographs

(see Plate I for anexample) or counted directly in disarticulated

and cleared specimens. Some ambiguity exists in counting the

number ofvertebrae towards the posterior end ofthe animal be-

cause the cartilaginous pelvic girdle and its attachment to the

backbone in particular, is not resolved in X-ray photographs.

Not surprisingly, both meristic characters proved highly cor-

related (cf. Romer & Parsons, 1977), with a product-moment

correlation coefficient of 0.70. Because neither data set was

complete,for three specimens the vertebrae count was estimated

from the regression equation of Ver versus Mio. Teeth counts

were not routinely performed, since this would require a partial

dissection of precious specimens. Instead we used the data of

Dolivo-Dobrovolsky (1926) after comparisonwith some of our

own specimens in order to confirm the reliability of the pub-

lished data.

At its 18th to 20th year, when Proteus matures and starts

reproducing, growth has not yet completely ceased (Durand &

Delay, 1981). By consequence, absolute morphometric values

are of limited use for comparative purposes. L was therefore

taken as a reference value for each of the continuous variables

in order to reduce the effect of variation in individual size. In

addition to a univariate character by character analysis using

analysis of variance, the "Wolterstorff Index" (WI = 100 x

LiE x Pa 1 (Wolterstorff, 1923; Arntzen & Wallis, 1993)) was

calculated. A Principal Component Analysis (PCA) including

a posteriori comparisons among localities was performed on

ln-transformed data. A further analysis was carried out using

Discriminant Analysis on apriorigroupedspecimens with locali-

ty as the independent variable. The statistical analyses were car-

ried out using the SYSTAT 5.1 software package (Wilkinson,

1989).

Blood, liver, heart, stomach, and a strip of muscle from the ven-

trolateral side (in Proteus) or from the tail (in Mertensiella and

Triturus, see below) were dissected from freshly sacrificed

animals. Erythrocytes were removed from the blood by brief

centrifugation. The plasma supernatant was diluted with an ali-

quot volume of 40% sucrose. Other tissues were ground in

homogenising buffer (0.1 M Tris, 10~ 3 M EDTA and 5 x 10~ 5

M NADP, adjusted to pH 7.0 with HCl) and centrifuged. The

aqueous supernatant was decanted and stored at -70 °C for

future electrophoresis.

Polyacrylamide slab gelelectrophoresis and staining of plasma


proteins were performed according to Maurer (1971). Enzyme

electrophoresis was performed using Connaught starch in hori-

zontal gels. The enzymes were visualised by standard histochem-

ical techniques (Shaw & Prasad, 1970; Harris & Hopkinson,

1976) with few modifications. Proteins assayed and buffer sys-

tems used are presented in Table I. Presumptive loci and alleles

(electromorphs) wereassigned numbers and letters, respectively,

when more than one zoneof activity was observed in sequence

starting from the most anodally migrating forms. For two en-

zyme systems, ACPH and GOT, two zones of activity were ob-

served, but only one could be consistently scored. These loci

were designated Acph-2 and Got-1, respectively.

Three different enzymes with protein digesting properties

were resolved. PEP-3, anodally migrating in the Proteus sam-

ples, showed substrate specificity for the tripeptide Leucyl-

Glycyl-Glycine, while the cathodally migrating PEP-1 and

PEP-2 were most clearly resolved when the dipeptideLeucyl-

Tyrosine was supplied as a substrate. Three plasma proteins

were scored, although in one unidentified protein (GP-2) the

corresponding electromorph(s) could not be identified for one

taxon. The fasteranodally migratingfractions were identified as

Albumin and Transferrin, based on their phenotypic charac-

teristics and their representation in the most concentrated frac-

tion in the plasma (cf. Rafinski & Arntzen, 1987).

To assess the degree of genetic variability within the Proteus

individuals, populations, and the outgroup taxa, we counted the

number of polymorphic loci and calculated the unbiased esti-

mate for meanheterozygosity based on Hardy-Weinberg expec-

tations (He) and the accompanying standard error (Nei, 1987).

In order to assess the degree of genetic differentiation across

populations we calculated the genetic distance measure of Nei

and its standard error (Z3Nej ; Nei, 1972), which may, for low and

intermediate values, be linearly correlated to time (Nei, 1987).

The matrix of DNcj

was converted into a phenogram using the

UPGMA method (Sneath & Sokal, 1973). A main objection

against this method is that the phenogram can be interpretedin

Table I. Electrophoretic conditions for 26 protein systems, correspondingto 40 loci, examined in the genera Proteus, Mertensiella, and

Starch buffers are A: Tris-citrate pH 6.0 (XIII); B: Tris-citrate pH 7.0 (I); C: Tris-citrate pH 8.0 (V); D: Lithiumhydroxide-tris-

citrate pH 8.1 (X); E: Tris-malate pH 7.4, electrode buffer is 0.22 M Tris, 0.10 M Maleic acid, 0.01 M EDTA and 0.01 M MgCl2 ,

gel

buffer is electrode buffer diluted at 1:10; F: Tris-EDTA-borate pH 8.9 (Ayala et al., 1972); G: Discontinuous tris-citrate-borate pH

8.2-8.7 (Poulik, 1957). Roman numerals refer tothe buffer systems of Shaw & Prasad (1970). PAGE refers to acrylamidegels according

to Maurer (1971: Table 4.1). Tissues used are: H = heart, L = liver, M= muscle, P

= plasma, and S = stomach.

Triturus.

Protein E.C. No. Locus Buffer Tissue

system extract

Acid phosphatase 3.1.3.2 Acph-2 A L

Adenosine deaminase 3.5.4.4 Ada C L

Albumin-

Alb PAGE P

Alcohol dehydrogenase 1.1.1.1 Adh-1, Adh-2 C L

Catalase 1.11.1.6 Cat D L

Esterase 3.1.1.1 Est-l, Est-2 D P,L

General Protein - GP-2 PAGE P

Glucose 6 phosphate dehydrogenase 1.1.1.49 G6pd-1, G6pd-2 C L

Glucose dehydrogenase 1.1.1.47 Gdh C L

Glucose phosphate isomerase 5.3.1.9 Gpi A L

Glutamate oxaloacetate transaminase 2.6.1.1 Got-1 G L

a-Glycerophosphate dehydrogenase 1.1.1.8 Gly-I, Gly-2 C L

Isocitrate dehydrogenase 1.1.1.42 Icd-1, Icd-2 E L

Lactate dehydrogenase 1.1.1.27 Ldh-1, Ldh-2 A,B H,M

Leucine aminopeptidase 3.4.11 Lap F L

Malate dehydrogenase 1.1.1.37 Mdh-1, Mdh-2 A L

Malic enzyme 1.1.1.40 Me C L

Mannose phosphate isomerase 5.3.1.8 Mpi- 1, Mpi-2 C L

NADH dehydrogenase 1.6.99.2 Nadhdh-1, Nadhdh-2 G L

Peptidase 3.4.11-13 Pep-1, Pep-2, Pep-3 D S

Phosphoglucomutase 2.7.5.1 Pgm-1, Pgm-2 A L

Phosphogluconatedehydrogenase 1.1.1.44 Pgd A L

Sorbitol dehydrogenase 1.1.1.14 Sdh A L

Superoxide dismutase 1.15.1.1 Sod-1, Sod-2 E L,M

Transferrin - Trf PAGE P

Xanthine dehydrogenase 1.1.1.204 Xdh-1, Xdh-2 E L

37Bijdragen tot de Dierkunde, 64 (I) - 1994

a phylogenetic sense only when rates of evolutionary change are

homogeneousacross the phyletic lines, a condition almost never

satisfied in reality. This strong and unreal assumption of unifor-

mity of evolutionary rates is relaxed in the distance-Wagner

procedure (Farris, 1972). This procedure requires a distance

measurewhich is metric. Since Nei's distance is not metric, we

applied the widely used Rogers' genetic distance DR (Rogers,

1972). From the resulting undirected trees the most parsimoni-

ous one, in terms of tree length, was selected. For an evolution-

ary interpretation the undirected trees have to be rooted. As

appropriately preserved specimens of the genus Necturus, the

assumed sister group ofProteus, were not available for compari-

son, instead seven Triturus cristatus (l.aurenti, 1768) from Can-

terbury and Peterborough, United Kingdom and two Merten-

siella caucasica (Waga, 1876) from near Rize in Turkey were

used as outgroups.

To distinguish on the generated tree the branches for which

the data show strong support from the areas in which the

branching order is more uncertain,we applied the jack-knife test

as advocated by Lanyon (1985) for distance data. The computer

programme employed for analysis was the micro-computer ver-

sion of BlOSYS-1 (Swofford & Selander, 1981).

The locality Jelsevnik was visited after heavy rainfalls, when

local inhabitants notified us about the wellingup of the so-called

"boiling-holes".To obtain an impression ofthe compositionof

the cave fauna that may serve as prey to Proteus,

drift nets of

various size and mesh width were placed immediately below the

holes as long as the discharge continued (generally from oneto

three days). The catches were generally very small and no

animals at all were obtained from the main Jelsevnik spring.

Since only empty gastropod shells could be collected from sedi-

ments of the outflow, it is not possible to establish their origin

precisely.

3 Results

3.1 Description of the black Proteus

Anticipating the discussion we assign taxonomie

status to the black Proteus that will be named Pro-

teus anguinus parkelj Sket & Arntzen.

Material. - Holotype: coll. nr. J8 is a female from Na Trati,

Jelsevnik near Crnomelj, Slovenia (plate I), preserved in ethanol.

Paratypes: same locality, six specimens (J1-J3, J6-J7, J9) and

one embryo/larva of 21 mm (J 10), preserved in ethanol. One

specimen cleared and stained (J5) and one disarticulated skele-

ton (J4) are preserved in glycerol. Others: one specimen (Dl)

from Doblicica spring, Doblice near Crnomelj, is ethanol pre-

served. All material is kept in the collection of the Department

of Biology of the University of Ljubljana, Slovenia, except for

J9, which is deposited in the Zoological Museum, Amsterdam,

the Netherlands and stored under nr. ZMA Herp. 9239.

3.1.1 Etymology. - "Parkelj" is one of the Devil's

names in Slovene Christian mythology. In Ljubl-

jana, around St. Nicholas day, small "parkeljni"

(plural of "parkelj") with a black body and a red

tongueare sold. These allegedly hypogean creatures

resemble theblack hypogean amphibian with its red

gills. Note that the ending "j" of the word parkelj

is not to be pronounced.

3.1.2 Diagnosis. - A dark pigmented and short-

snouted subspecies of P. anguinus with externally

visible eyes, a head with angular and convex lateral

sides, a long trunk, short legs, and a short tail.

3.1.3 Description of the type series. - Holotype:

morphometric characters, variability indices, and

corresponding data for white Proteus a. anguinus

can be taken from Table II. The dorsal and lateral

parts of the body are almost completely black with

a violet or brownish hue. In some specimens the

colour is only very dark brown. The snout is broad-

ly bordered black as soot, in some specimens with

a pale or whitish triangle in the preocular region.

Limbs pale with a dark patterning. Ventral side of

the body pale with a bluish or pinky shade. Skin

surface finely granulated in preserved specimens.

An alcohol preserved specimen (Dl) has retained a

dark brownish black coloration for seven years.

Gills red in live specimens, with dark brown prin-

cipal branches.

The skin contains large quantities of pigment,

mainly in the upper part of the dermis. The upper

dermis also contains more numerous multicellular

glands than is found in white specimens and Leydig

cells are present in the epidermis (Bulog, 1991,

1992).

The eyes are small and externally visible as black

dots that are surrounded by a white circle. The eye

is covered by a distinct transparent conjunctiva but

without lids or similar structures. The diameter of

the conjunctiva is ca. 5% of the head length. The

general structure of its bulb is similar to that in

epigean amphibians with a well-developed retina.

In a 190 mm long specimen from Doblicica, the

bulb diameterwas measured to 1.3 mm (7% of Lc)

and the diameterof its lens 0.2 mm (glutaraldehyde

preparations for electron microscopy were made by


Bulog (1991, 1992)). In a skinned specimen of 205

mm, at 0.9 mm the eye bulb's diameter was 3.9%

of Lc (ethanol preparation).

The head is slightly broader than the trunk and

approximately 11 % of the length of the body (L-

Led). In its postocular part the head is parallel-

sided, the short snout shaped as a trapezoid with

rounded corners. The snout comprises on average

27% of Lc. Three pairs of subdermal muscular

cushions (that we tentatively identify as the leva-

*Holotype.

from Planina, Rupnica,

and Vir. L = total length; Lc = head length; Ltc = head width; LaP = pre-pedal length; LiE = trunk length; PaL = anterior leg

length; PpL = posterior leg length; Lcd = tail length;Myo = number of trunk myomeres; Ver = number ofneck and trunk vertebrae;

WI = Wolterstorff Index. Values in parentheses are estimated from the regression of Led versus L and Ver versus Myo. Data are pre-

sented individuallyfor all biochemically studied specimens and for the type series from Jelševnik. Statistical data include the specimens

measured by Kranjec (1981).

Table II. Morphometric data (mm) for Proteus anguinusparkelj spp. n. from Jelševnik and P. a. anguinus

Locality Specimen

number

L Lc Ltc LaP LiE PaL PpL Led Myo Ver WI

JeUevnik

Jl 227 27 16 31 130 14 13 64 31 (33) 10.8

J2 247 26 16 32 138 14 12 (69) - - 10.1

J3 240 25 17 34 140 16 13 67 32 35 11.4

J5 205 23 14 31 120 13 11 55 - 34 10.8

.17 199 20 12 27 113 13 10 57 30 (33) 11.5

J8* 276 28 17 37 158 16 13 78 30 34 10.1

J9 232 24 16 30 135 14 13 65 33 35 10.4

sample size 7 7 7 7 7 7 7 6 5 4 7

mean 232.3 24.7 15.4 31.7 133.4 14.3 12.1 64.3 31.2 34.5 10.7

SD 26.0 2.69 1.81 3.15 14.6 1.25 1.22 8.19 1.30 0.58 0.57

minimum 199 20 12 27 113 13 10 55 30 34 10.1

maximum 276 28 17 37 158 16 13 78 33 35 11.5

Planina

PS 235 26 17 35 118 19 16 75 27 31 16.1

P6 255 30 17 37 129 19 20 82 28 31 14.7

sample size 54 54 54 54 54 54 54 54 54 53 54

mean 223.2 29.0 15.3 36.0 109.9 18.7 17.5 72.6 25.5 30.8 17.1

SD 26.6 3.47 2.40 4.02 14.1 2.07 2.15 9.29 0.79 0.51 1.20

minimum 147 19 9 24 72 13 12 48 24 30 13.8

maximum 299 37 23 46 152 24 23 100 28 32 19.6

Rupnica

Rl 270 31 19 42 133 25 21 90 25 29 18.8

R2 225 24 16 31 111 18 15 77 26 31 16.2

sample size 7 7 7 1 7 7 7 7 7 7 7

mean 245.0 27.7 17.5 36.6 118.4 20.4 17.2 83.9 25.4 29.9 17.3

SD 40.6 4.82 3.37 6.45 21.1 3.40 2.53 11.7 0.54 0.69 1.64

minimum 188 23 14 29 88 16 15 67 25 29 15.7

maximum 294 34 22 45 144 25 21 102 26 31 19.7

Vir

V7 219 24 15 32 108 19 17 76 27 31 11.1

sample size 14 14 14 14 14 14 14 14 14 14 14

mean 228.4 27.8 16.0 34.6 111.4 19.1 16.7 76.5 26.1 30.6 16.7

SD 34.7 3.90 2.61 4.55 16.9 3.08 2.00 12.8 0.66 0.65 1.95

minimum 154 21 11 26 76 14 14 49 25 29 11.1

maximum 285 34 20 41 138 23 20 100 27 31 18.7


tores mandibulaeanteriores, levatores mandibulae

externi, and depressores mandibulae posteriores)

give the head a bulky appearance. Three pairs of

short, bushy gills are well developed with a length

of approximately 20% of Lc.

The trunk is cylindrical and anguilliform, on

average 58% of the length of the body. The flanks

of the body are running parallel with shallow fur-

rows between the myomeres. The tail is short, com-

prising 28% of the entirebody length and flattened

laterally with the ventral and dorsal sides running

almost parallel. The tail is as high or slightly higher

than the trunk. The tail tip is broadly rounded or

triangular with a rounded tip. The legs are short

and slender. The anterior legs form on average 6%

of the body length and have three toes. The poste-

rior legs are slightly shorter and have two toes.

The skull is similarly built to that of white speci-

mens (cf. Dolivo-Dobrovolsky, 1926) but propor-

tioneddifferently. The neurocraniumis wider, with

a rostro-occipital length that is slightly less thanthe

length of four pectoral vertebrae. The dentalia are

placed in a wide arch, the largest width equals the

lengths of 2.3 pectoral vertebrae. The ratio rostro-

occipital length versus largest mandibular width

averages 1.4 to 1. The praemaxillaria are less than

twice as long as wide (Fig. 2). Theethmoidal plaque

in the only macerated skull that we possess has a

rather peculiar shape with its rostral tines curved

and with convergent tips (Fig. 3). In the frontalia

the width versus length ratio averages at 0.6 to 1.

The frontaliaslightly protrude along the praemaxil-

laria. The other bones of the neurocranium are

Fig. 2. Cranial and mandibular bones in Proteus anguinus

parkelj from Jelševnik (specimen no. J4): a, angulare; d, den-

tale; f, frontale; o, operculare; p, parietale; pm, praemaxillare;

pq, paraquadratum; q, quadratum. Scale bar 5 mm.

Fig. 3. Rostral part of the trabecula cranii with the anterior

ossification and ethmoidal plaque in Proteus anguinusparkelj

from Jelševnik (specimen no. J4). Scale bar 1 mm.


Number of teeth on

Praemaxillare Vomer Pterygo- Dentale Operculare

palatinum

P. a. anguinus>*

(minimum)

P. a. anguinusj*

(maximum)

P. a. parkelj

P. a. parkelj

left 7 22 5 22 6

right 7 22 6 21 5

left 10 or 8 (2) 24 (8) 6 33 or 23 (11) 12

right 9 (2) or 8 (4) 25 (8) 6 31 (2) or 23 (8) 12

J4—left 5 19 4 19 7

J4—right 5 18 4 16 ?

J5—left 7 20 5 17 (6) 8

J5-right 8 22 5 17 (4) 6

short and wide. There are fewer teeth than in the

white morph, particularly so on the dentale. On the

operculare, teeth are arranged in two or three rows.

The dentale bears 16 to 19 teeth with another 4 to

6 small auxiliary teeth that are placed outside the

main row (Table III).

The number of vertebrae between the skull and

the posterior legs is 34 to 35. The neck vertebraeare

of the same length as the pectoral ones. The caudal

vertebrae get gradually smaller till number XXIII

or XXIV and are then followed by 3 to 5 rudimen-

tary vertebrae that have no processes. Up to num-

ber XVII, the neural spine of the vertebrae is in the

shape of a sharp and caudally extended triangle.

The posterior spines form irregular trapezoidal

crests and owing to the fact that they are remark-

ably short, the vertebrae appear to be high.

A drifted larva of 21 mm L (specimen no. J10)

was in Briegleb's developmental stage 21 (Fig. 4). It

did not have the appearance of an active feeder. Its

anterior legs are short and three-toed, the posterior

ones are only short stumps without toes. The eye di-

ameter is 8% of Lc. Ventrally the specimen is pale

white. Dorsally it is densely speckled grey and less

so on the sides.

As a result of the allozyme study, four alleles,

Catb

,Cat

d,

Nadhdh-la

,and Pgd

a

,were uniquely

observed in black specimens.

3.2 Comparison with related taxa

3.2.1 General appearance, pigmentation, eyes, and

gills. — Differently from the new subspecies, all

Proteus from other populations than Jelsevnik and

Dobliöica exhibit a nearly white or pinkish skin,

sometimes with pale yellow or pale grey patches.

They only become dark after a long exposure to

light.

In the stages immediately before or after the lar-

val hatching (corresponding to an advanced stage

21 of Briegleb) the eye, with a diameterof 6 to 8°7o

of Lc, is not evidently larger in P. a. parkelj than

in P. a. anguinus (corresponding value is 8%; mea-

* According to Dolivo-Dobrovolsky (1926).

Fig. 4. Anterior part of embryonic larva of Proteus anguinus

parkelj from Jelševnik (specimen no. J10) at Briegleb stage 21.

Scale bar 1 mm.

Table III. Number of teeth observed in P. a. anguinus (four specimens from Luce and Stična, according to Dolivo-Dobrovolsky, 1926)

and in two specimens ofP. anguinusparkelj ssp. n. from Jelševnik. Figures in parentheses denote the accessory smaller teeth standing

outside the main row. Maximal numbers for white specimens are given separately for different combinations of principal (large) teeth

and accessory teeth. In both taxa the opercular teeth may be arranged either in one or in two to three rows.

Bijdragen tot de Dierkunde, 64 (1) - 1994 41

surements taken from a drawing in Briegleb (1962)

and from photographs by Vandel & Bouillon (1959)

and Vandel et al. (1966)). The dark pigmentation in

the embryo of P. a. parkelj seems to be only a little

denser than in P. a. anguinus. The intense pigmen-

tationof adult P. a. parkelj differs from most sur-

face amphibians by the absence of a more vivid

pattern of coloration. Also in the genus Necturus,

sister group of Proteus, most species are without

any bright coloration (cf. Conant, 1975; Behler &

Wayne King, 1979).

The eye in the black morph is small and without

lids and superficially similar to that in Necturus. In

the white morph the degenerated eyes are always

covered by skin, although in some cases the eye

might be shining through as a blackish dot. In a

skinned specimen of white Proteus, the eye bulb's

diameter at 2°7o of Lc was half that of the black

specimen. As inother species with rudimentary eyes

(Peters et al., 1973; Sket, 1985) the diameterof the

rudiment is expected to be highly variable.

The length and ramification of the gills is vari-

able within white populations. It may be that the

varying oxygen concentration in the water affects

increase or decrease of gill size. A similar phenome-

non has been supposed for Necturus (Conant,

1975). Compared to other Proteus populations, the

gills in the black morph are of medium size.

3.2.2 Univariate morphometric and osteological

comparisons. — The total length of the black Pro-

teus ranged from 199 mm to 276 mm, which falls

within the range observed in our sample of white

Proteus (range 147-299 mm).

For most measurements relative to L (Lc, LaP,

LiE, Lpa, Led) differences are found between

populations. RelativeLc is smaller in theblack Pro-

teus than in the whites, while relative Ltc of both

forms is nearly equal. The head differs strikingly in

shape, more strongly than is brought about by the

morphometric data. In theblack morph the head is

parallel-sided behind the snout. In the whites, due

to a weaker musculaturebehind theeyes, the whole

head has the shape of a long, truncated triangle

(Fig. 5). It even can be pear-shaped with concave

flanks.

The relative length of the posterior legs, strongly

correlated with relative head length (product-

moment correlation coefficient = 0.94), is smaller

in the black morph than in the white morph. Mak-

ing up on average 58% of L, relative trunk length

is larger in the black morph than in the white morph

in which the corresponding value is 49%. Cor-

respondingly, in the black morph the absolute num-

ber of trunk vertebrae and myomeres is 4; 5 or 6

higher than in the white morph.

Both pairs of legs are remarkably shorter in the

black specimens than in the white ones. In both

morphs the legs are similarly built with only three

and two toes at the foreand hind limb, respectively,

and with a simplified skeleton (cf. Aljancic & Sket,

1964).

The tail is longer in white Proteus than in black

Proteus. The tail shape shows a great interpopula-

tional variability but some contour characters are

quite constant within populations. In specimens

from Planina, the tail with its crest is of the same

height as the body, gradually lowering towards the

tail tip, which is usually bluntly pointed. In speci-

mens from Rupnica and Vir the tail is usually higher

than the body. It does not narrow caudally and ends

broadly rounded. In the black morph the tail tip is

Fig. 5. Skinned heads ofProteus anguinusanguinus from Plani-

na (a) and P. a. parkelj from Jelševnik (b).


of a shape somewhere in between the recorded ex-

tremes.

In the skull of the black morph, the flat bones

such as the parietalia, frontalia, and praemaxillaria

are shorter or wider than in the white morph and the

dorsally visible intercalations are shorter. Also some

other bones (dentalia, paraquadrata) are differ-

ently shaped. The mandibular arch such as formed

by the dentalia and angularia is wider than in the

white morph. The number of teeth (Table III) is

higher in the white morph than in theblack morph,

the difference being most pronounced in the den-

tale. The fossil relative Mioproteus caucasicusEstes

& Darevski with 10—12 teeth on its vomer (Estes &

Darevski, 1977) was markedly less toothed.

Some significant morphometric differences are also

found between white Proteus populations from

Planina and Sticna. The two Sticna populations

(Rupnica and Vir), however, are mostly indistin-

guishable.

In some white specimens the neck vertebrae are

elongated, a feature that has not been observed in

theblack morph. Otherwise the trunk vertebrae are

similarly shaped, and even in the black morph the

vertebrae are not as stout as in Mioproteus caucasi-

cus (cf. Estes & Darevski, 1977; Estes & Schleich,

1993). In Necturus as well as in all Proteus the cen-

tra and particularly the processes of the tail ver-

tebrae are less well developed than in Paleoproteus

(cf. Herre, 1935) or in neotenicand regular Triturus

alpestris. In Proteus the centre of the individual

vertebra is thinner, the dorsal spine and ventral

process are narrower and the interspaces between

subsequent vertebrae are substantially larger. Alto-

gether their shape is variable. In black specimens,

the caudal vertebrae get gradually smaller till the

XXIIIrd or XXIVth and are followed by three to

five rudimentary ones which are without processes

(Fig. 6). In the white specimens from Vir the tail

vertebrae are similar to the inspected black speci-

mens, although markedly shorter. In the white

specimens fromPlanina the vertebraeare lower and

gradually diminishing in size towards the tip of the

tail. One specimen has a chainof seven small, spine-

less vertebrae after the XXVIIIth that still bears

spines (Fig. 6), while there are only three of such in

another specimen.

Fig. 6. Last caudal vertebrae in Proteus anguinusparkelj from Jelševnik (a) and P. a. anguinus from Vir (b) and Planina (c). The XVIIIth

caudal vertebra is indicated by an asterisk. Scale bar is 5 mm. The vertebrae are seen here from the same perspective as the heads in

Fig. 5.


3.2.3 Bivariate and multivariate morphometric

analysis. — The Wolterstorff Index is significantly

different in the black and white morph of Proteus

(P< 0.001; Table II). Similar to the situation in

Triturus cristatus for which the WI was originally

devised (Wolterstorff, 1923), the number of trunk

vertebrae(Ver) is an equally good or better discrim-

inatory feature than WI itself (Arntzen & Wallis,

1993) (Fig. 7ab).

In the multivariate PCA all eight variables have

high component loadings on the first axis, which is

interpreted as representing general size. In line with

the univariate analysis, size is not population de-

pendent and not discriminatory for black versus

white populations. The second axis is dominatedby

LiE, PaL, and PpL. Component loadings have

contrasting signs, which means that animals have

either a short trunk and long extremities, or a long

trunk and short extremities. Hence the second PC A

axis is essentially the same as the WI (Fig. 7bc).

On the third axis Lc and PpL dominatetogether

with Led, with contrasting signs. This axis helps

separating all populations from one another, espe-

cially those from Planinaand Jelsevnik from Rup-

nica and Vir. Plotting the second versus the third

axis provides excellent overall separation of black

and white specimens (Fig. 7c).

Discriminant analysis reveals that the morpho-

metric differentiationfor Planina versus Rupnica

and Vir is not statistically significant, with 8 out of

54 specimens (15%) wrongly classified. The popu-

lations of Rupnica and Vir are morphometrically

virtually indistinguishable, with 7 out of 21 speci-

mens (33°7o) wrongly classified.

3.2.4 Protein analysis. - In 23 different enzyme

systems and three plasma proteins the variability

could be consistently scored, with the exception of

some loci in the outgroup taxa (two loci and four

loci in Mertensiella and Triturus, respectively).

Data on allelic composition for altogether 40 pre-

sumptive loci and the associated estimates of ge-

netic variability of populations are presented in

Table IV. The Mpi-J locus showed no variability

over any of the assayed populations, whilean addi-

tional 11 loci showed no variation across Proteus

samples. In 12 loci, inter-but no intra-populational

variation was observed for Proteus populations.

The remaining 16 loci showed variability in at least

oneof the Proteus samples. The mean heterozygos-

ity (He) for the combined Proteus populations is

12.0 ± 4.0%, which is similar to the value found

for M. caucasica (11.2 ± 3.3%) and substantially

higher than He

= 1.3 ± 1.0% found for T. crista-

tus. Among Proteus populations He

ranges from

7.5% for the two studied specimens from Planina

to 22.5 ± 6.7% for the single individual from Vir.

Fig. 7. Morphometries of Proteus anguinus anguinus (outline

symbols) and P. a.parkelj (solid symbols). Upper part (a): uni-

variate - number oftrunk vertebrae (Ver); to the middle (b): bi-

variate - Wolterstorff Index (WI, details see text); lower part

(c): second and third axes of a Principal Component Analysis

(PCA) with centroids and convexoutline polygons for P = Pla-

nina, R = Rupnica, and V = Vir (P. a. anguinus) and J =

Jelševnik (P. a. parkelj).


Species

Locality

Reference no.

Proteus

anguinus

Jelsevnik, Planina, Rupnica, Vir,

Slovenia Slovenia Slovenia Slovenia

J1 J2 P5 P6 R1 R2 V7

Mertensiella

caucasica

Azaklihoca Köyü,

Rize, Turkey

ZMA Herp.

7277 (n =2)

Triturus

cristatus

Canterbury and

Peterborough, U.K.

ZMA Herp. 9199

and 9200 (n =7)

Individuals from two populationsofthe latter taxon are pooled. Allelic compositionin parentheses

refers tosingle individuals. Missing data are indicated by a hyphen. P is the number ofpolymorphic loci, He

is the average heterozygos-

ity calculated on the basis of Hardy-Weinberg expectation and SE is the corresponding standard error (in percent values).

Triturus.andProteus, Mertensiella,

Table IV. Allelic composition for 40 gene loci (upper part) and measures of genetic variability (lower part) as observed in the genera


High values of Z?Nej (> 1.8; Fig. 8a) were ob-

tained for all comparisons among Proteus and the

outgroup taxa. Among Proteus populations rela-

tively high values, with Z)Nei ranging from 0.41 —

0.56, were obtained for comparisons involving the

Planinapopulation versus theothers. Substantially

lower values (0Nei= 0.23 and 0.24) were recorded

for comparisons between the black population

from Jelsevnik and the white populations from

Rupnica and Vir, respectively. In all cases the stan-

dard error of these estimates, averaging at 32% of

£>Nej,

is substantial and taking the standard error

into account, the latter populations are not clearly

separated (Fig. 8a). Using the UPGMA-method

with the matrix of Z?Nei

results in a tree in which,

seen from the root, the population from Planina

branches off first, followed by Jelsevnik and the

two remaining clustering populations of Rupnica

and Vir (Fig. 8a). Each of the jack-knife replicates

supports this branching topology.

In the distance-Wagner tree that displays maxi-

mum parsimony theresults are essentially the same.

Inspecting the distribution of alleles, seven alleles

(Adh-2d

,Alb

b

,G6pd-P, G6pd-2

b

,Icd-2

b

,Mdh-2d

,

and Sod-2c) are potentially synapomorphic charac-

ter states defining the grouping together — separate

from Planina — of the Jelsevnik and both Sticna

populations. Conflicting results and hence no ro-

bust phylogenetic resolution is found within the lat-

ter group (Fig. 8b).

Four alleles, Catb

,Cat

A

,Nadhdh-la

,and Pgd a

,

were observed uniquely in the black Proteus from

Jelsevnik. Larger reference samples are required to

find out whether or not these alleles are strictly aut-

apomorphic. No loci with fixed alleles diagnostic

for the black phenotype were observed. The popu-

lations from Rupnica and Vir (ö Nej= 0.14) differ

from one another most markedly on both Ldh loci.

3.3 Ecology and reproduction of the black Proteus

3.3.1 Habitat: abiotic parameters. — Up to the

present, P. a. parkelj has only been documented for

two localities near the town of Crnomelj in the Bela

Krajina, which is in the southeasternmost part of

Slovenia. The Doblicica and Jelsevnik springs both

discharge into the river Doblicica. They are situated

Fig. 8. To the left (a): UPGMA-dendrogram for four populations of Proteus (white morph with “outline” printed population name)

based upon Nei’s genetic distance as measured over 40 gene loci. Boxes indicate estimated standard error (Nei, 1987). The cophenetic

correlation coefficient is 0.96. To the right (b): minimum length distance-Wagnertree on the basis of Rogers’ genetic distance, rooted

by the outgroup. Length of the tree is 0.62 (1.75 with outgroup taxa included); the cophenetic correlation coefficient is 0.99. The inter-

rupted line refers to a branch that is not robust under the jack-knife test (see text). Outgroup taxa Mertensieila and Triturus are included

for comparison.

46 B. Sket & J. W. Arntzen - A black Proteus

in close proximity to one another (about 2.5 km

apart), below the higher conical karst of the Pol-

janska Gora mountainridge, with deeply set water

channels. At an altitude of 150 m above sea level,

the conical karst runs out in karst plains (Habic,

1991). On the plain, some small rivers flow on the

surface while others are subterranean at a depth of

several meters.

Doblicica is a permanent spring, rising from a

deep pool of ca. 25 m diameter. There are some

5-11 m deep pits in the pool and its outflow. One

black Proteus was caught in theoutflow (not in the

main spring) after a pumping experiment, which

lowered the water level by two meters (Aljancic

et al., 1986). Soon after, oneof us (BS) inspected the

spring by diving, but no amphibians were seen. At

Jelsevnik, there is a permanent limnocrene spring

(called Jezero), similar to the Doblicica spring, and

two associated groups of temporarily active "boil-

ing holes", all within a reach of 150 m. One black

specimen was photographed below the southern

group of boiling holes (Jamnice), while the typical

series was caught at the northern group of holes of

1 to 100 cm in diameter (Na Trati). The boiling

holes eject water only a few times a yearafter heavy

rains. When they are active or soon afterwards,

black specimens have occasionally been found

crawling around on the wet meadow or they were

collected straight from a narrow hole in the

meadow covering the karstified rock. The boiling

holes at Jelsevnik, as well as some accessory springs

in the outflow of Doblicica, are probably fed by

shallow local aquifers from the mountain or from

the low karst zone. When active, they discharge

large amounts of soil and some terrestrial animals.

The presence of some epigean aquatic organisms in

the drift of Na Trati suggests its partial origin from

a surface stream in the background.

Physical and chemical parameters of the perma-

nent springs were measured by Habic et al. (1990)

on 17 occasions in spring and autumn of 1986—

1990. In the permanent springs of Doblicica and

Jelsevnik, water temperature fluctuated between

10.0 and 11.3 °C and temperatures in winter and

summer are expected to deviate only marginallyfrom this range (the extreme value of 14.8 °C as

mentioned for Doblicica is almost certainly errone-

ous). The total hardness of the water ranged from

200 to 250 mg CaC03

x l"1

,with Ca x mg~' ra-

tios fluctuating. Oxygen content in Doblicica in the

autumn of 1986 was near saturation. In May 1987,

O-phosphates in Doblicica and Jelsevnik springs

corresponded to 0.01 mg PO x 1_1.

The dissimilar

response of the springs to large scale water extrac-

tion and their differing chemicalproperties suggest

complex and perhaps remote connections between

these springs and the areas from where the water

originates (HabiC et al., 1990).

The hydrological complexity of the region can be

illustrated by the situation at Jelsevnik on Novem-

ber 1, 1992. While the temperature of the Jezero

permanent spring and some of the boiling holes was

in the range of 10.1 to 10.3 °C, most holes dis-

charged water of 11.2 °C, which was also the tem-

perature of a nearby epigean stream. Close moni-

toring of the system, especially during upwelling

periods, will be required to gain a proper under-

standing of the situation.

3.3.2 Associated biocoenoses and food. - Some

stygobiont animals have been ejected from the per-

manent Doblicica and Jelsevnik springs (Table V).

Since Na Trati has no steady outflow and associat-

ed surface fauna, the surface animals in its drift

either originate from an unidentified stream, or

they are sucked in hydrodynamically from the main

spring.

Upon capture, some of the Proteus gave up their

stomach contents. Surprisingly, food items pre-

dominantly consisted of surface-dwelling prey spe-

cies. Of these, the gastropod Vitrea sp., earth-

worms (Oligochaeta: Lumbricidae), and the chilo-

pod Lithobius sp. may have actively penetrated

subterraneously, but the aquatic gastropod Sadle-

riana fluminensis Kuester has been shown not to

enter cave waters (Sket, 1970). These observations

are in favourof suggestions that Proteus forages at

night out in the open. In captivity, black Proteus

are seen to be active and aggressive feeders that

may, at least in an illuminated aquarium, attack

and bite theirwhite counterparts. Indeed, a violent

attack on a white specimen during feeding has been

registered on videotape.


cd

Jelsevnik>o

3

QJ Ja NI N2

Stygobiont and stygophilic species

Gastropoda

Belgrandiellafontinalis ( Schmidt, 1847) x x x x

Belgrandiellasp. x x x x

Hauffenia cf. michleri Kuscer, 1932 x x

Hauffenia sp. x x x x x

Iglica cf. hauffeni (Brusina, 1885) x

Iglica sp. x

Sadleriana cavernosa Radoraan, 1978 x x x

Sadleriana sp. x x x

Crustacea

Microcharon sp. x

Monolistra racovitzai spp. Strouhal,

1928 x x

Monolistra velkovrhi Sket, 1960 x x

Niphargus sp. x x

Niphargus sp. x

Niphargus sp. x

Niphargus subtypicus Sket, 1960 x

Proasellus parvulus (Sket, 1960) x

Troglocaris sp. (larva) x x

Troglodiaptomussketi Petkovski, 1978 x

Surface aquatic species

Cyclopoida x x x

Cyclostomata (larva) x

Gammarus fossarum Koch, 1836 x

Hydra sp. x

Pisces (larvae) x

Terrestrial species

Chilopoda x x

Diplopoda x x

Diplura x

3.3.3 Reproduction. - The finding at Jelsevnik of

an embryonic larva of stage 21 (still swollen by

vitellus and probably not able to move around in

an organised way) speaks in favour of oviparity in

P. a. parkelj, as is likely to be the case in P. a. an-

guinus (Briegleb, 1962; Sket & Velkovrh, 1978).

4 Discussion

4.1 Taxonomic status of the black Proteus

The black morph of Proteus can be distinguished

from the white morph by: (1) synthesis of pigment

in the absence of light, (2) fully developed eyes, (3)

the shape of some cranial bones and the numberof

teeth, (4) a differenthead shape stemming from a

well-developed head musculature, (5) a high num-

ber of trunk vertebrae and its associated long trunk,

(6) short relative length of the tail, and (7) short

relative length of the legs.

On the basis of these evidently constant differ-

ences the black and white morph deserve attribu-

tion to separate taxa. On many occasions differ-

ences of similar magnitude have led to the recogni-

tion of separate species or even genera. Examples

include the characinid fish Astyanax fasciatus

Cuvier, 1819 and its cave relative Anoptichthys jor-

dani Hubbs & Innes, 1936, that have recently been

shown to be completely interfertile(Culver, 1982).

Unfortunately, in Proteus crossing experiments

pose the greatestof technical difficultiesand to test

for interfertility will be almost impossible. So far

both morphs have in nature only been documented

in allopatric conditions. The observed genetic dis-

tance of Z)Nei

= 0.24 between the black Proteus

and the white population from Sticna is similar to

the distances found for the various forms within the

Triturus cristatus superspecies, that to a large

extent are genetically isolated from one another

(Wallis & Arntzen, 1989; Arntzen, in prep.).

According to the results of the allozyme study,

the black morph is phylogenetically closer to the

populations of Sticna than to the population of Pla-

nina. In view of them being morphologically simi-

lar, the large genetic distance between Planina and

Sticna populations is remarkable. If a study of

populations that are geographically intermediate

would fail to indicate a pattern of clinal variation,

we would identify the populations of Sticna as a

species separate fromP. anguinus. Nomenclatorial

priority would go to Proteus zoisii (Fitzinger, 1850)

of which Rupnica is the type locality. The morpho-

logically indistinguishable and biochemically simi-

lar population from Vir should be attributed to the

Table V. Fauna drifted from the habitat of Proteus anguinus

parkelj: J, Jelševnik Jezero; Ja, Jamnice; N1, Na Trati (big

hole); N2, Na Trati (small holes). Gastropods were collected

from sediments. Taxa of which the origin is difficult to deter-

mine, such as Nematoda, Oligochaeta, and Cyclopoida, are ex-

cluded (for details see text).

48/I B. Sket & J. W. Arntzen - A black Proteus

P. a. anguinusandP. a. parkelj in lateral view (pho-

tos: A. Hodalic).

Proteus anguinus parkelj from

Jelševnik in dorsal view, the insert shows the head in de-

tail (photo: B. Sket); middle and bottom: anterior parts

of

from left to the right specimen numbers are

J9, J8, J3, R1, and P5 (see Table II; photo: V. Tabor);

left: live specimen of

andP. a. parkelj P.

a. anguinus,

Plate I. Top: X-ray photographsof


same taxon (although previously it has been given

a name of its own: Hypochthon schreibersii

Fitzinger, 1850). If such a revision would be fol-

lowed, the black morph would be attributed to

P. zoisii as the subspecies P. z. parkelj. An osteo-

logical synapomorphy in the shape of a less equally

segmented tail skeleton, such as found in the popu-

lations of Jelsevnik and Vir (Fig. 6) might speak in

favour of a taxonomie separation of the Planina

population.

Recognition of two species of Proteus would

raise the question whether or not the black morph

would also deserve specific status. A genetic dis-

tance of £>Nei

= 0.24 is not a priori incompatible

with this view, while the morphological distinctive-

ness of the black morph speaks in its favour. The

finding of genetically isolated sympatric black and

white Proteus would help proving the point. In the

absence of firm phylogenetic, ecological, and ge-

netic datawe as yet refute such taxonomie revisions

in favour of conservatively recognising a single spe-

cies rather than two or three species.

4.2 Biogeographic and evolutionary history

At present the regions of Planina, Sticna, and

Jelsevnik drain towards the Sava and the Danube,

but little doubt exists about the mutual isolation of

their hypogean habitats (cf. Gams, 1965; Novak,

1990). The genetic data suggest that the western

Proteus population from Planina has become iso-

latedfrom the eastern Sticnaand Jelsevnik popula-

tions earlier than the Sticna and Jelsevnik popula-

tions with regard to each other.

In amphibians the "molecular clock" has been

calibrated as one unit of Z)Nei

reflecting 14 My. of

lineage separation (Maxson & Maxson, 1979). For

the genus of aquatic salamanders Triturus it has

been shown that this calibration fits with indepen-

dently derived estimates of speciation times; on that

basis a consistent biogeographic scenario could be

put forward, which corresponds to the vicariance

biogeography of other amphibian and non-amphi-

bian taxa (Oosterbroek & Arntzen, 1992). Applying

this calibration to Proteus, while keeping the stan-

dard error to the estimate of genetic distance in

mind, this would mean that the first documented

separation of Proteus lineages took place some 9-5

My. ago, which corresponds to the upper Miocene

or lower Pliocene. Following the same line of argu-

mentation, the Sticna and Jelsevnik populations

show lineage independence for 4.5-1.1 My. This

coincides with the upper Pliocene to the lower Pleis-

tocene.

It is generally agreed that the prekarstic Miocene-

Pliocene surface streams of the Sticna region and

the Bela Krajina flew in approximately the same

direction as they do today. There is, however,

some controversy concerning the situation around

Postojna (cf. Melik, 1951, 1952; Jenko, 1959).

Some Zoogeographie data, including the distribu-

tion of the isopod crustaceans Asellus aquaticus

cavernicolus Racovitza, 1925 and Proasellus istria-

nus (Stammer, 1932) speak in favour of an ancient

hydrographie connection between the Postojna

region and the Gulf of Trieste (Sket, in prep.). At

the end of the Miocene, southern Slovenia became

tectonically active which resulted in the presence of

large depressions with lakes (Prelogovic et al.,

1975) and in changes in the directionof the flow of

rivers. We suggest that Proteus may have formed a

series of taxa in surface rivers and lakes before the

Pliocene, with morphological differentiation de-

veloping between them. Proteus appears to be an-

other example of a cave species in which the dis-

tribution and phylogenetic relationships relate to

the prekarstic paleohydrology, rather than to the

present-day situation (cf. Sket, 1970, 1986; Sket &

Bole, 1982).

With the karstification, a new, hypogean habitat

became available to Proteus. Governed by environ-

mental conditions, the phenotypical differentiation

may locally have become obscured through evo-

lutionary convergent, troglomorphic adaptations.

Why the black Proteus has not undergone such

adaptive change remains an open question. The

retentionof non-troglomorphic traits may be asso-

ciatedto weak selection (cf. Sket, 1985) or to a rela-

tively recent colonization of the hypogean habitat.

The Pleistocene has seen several glaciations but

no evidence is available to document that these

extended onto Slovene territory before 1 My. ago

(cf. Bowen et al., 1986; V. Pohar, pers. comm.).


Intensive karstification, moving surface streams

underground, caused complex hydrographical

changes. With the karstification proceeding, the

Proteus populations from Sticnaand Jelsevnik may

have become isolated from one another, but not

necessarily in caves. In a time span similar to their

isolation or less, some stocks within the genus

Triturus developed into clearly distinct species

(Rafinski& Arntzen, 1987; Oosterbroek & Arntzen,

1992).

4.3 Character evolution

WithinProteus, thepresence of externally differen-

tiatedeyes seems unique to P. a. parkelj. Kammerer

(1912), on the basis of experiments that so far have

not successfully been repeated (Durand, 1973),

claimed that under the influenceof light, blind Pro-

teus may develop eyes and pigmentation. Occasion-

al observations on captive Proteus do suggest that

the ability of pigment synthesis is retained indeed

(Ehrenberg, 1867 in: Knauer, 1878), but thoroughly

conducted experiments have not been carried out in

Proteus, nor in any other cave amphibian. The

newly developed pigmentation may cover the whole

body or only part of it (AljanCiC et al., 1986), while

retaining a white triangle on the snout (Ehrenberg,

1867 in: Knauer, 1878) - perhaps not unlike that

found in P. a. parkelj.

The Jelsevnik population of Proteus is pigment-

ed although specimens are normally not exposed to

daylight. As in Proteus,theatavism of melaninsyn-

thesis is found in some cavernicolous fish while

others seem to have completely lost the ability (see

e.g. Vandel, 1965: 408). The pigmentation of P.a.

parkelj,„, along with the well-developed eyes, are

considered to be plesiomorphic traits. Since the

duck-bill shape of the P. a. anguinus head is an at-

tribute typical of troglobiont vertebrates (cf. Van-

del, 1965; Cooper & Kühne, 1974), occurring in

fishes as well as in plethodontid amphibians, it can

be regarded apomorphic compared to the plesio-

morphic shape of P. a. parkelj. The change in head

shape seems to be associated with a narrowing of

the neurocranial bones and the mandibular arch

and a weakening of the major head musculature.

The elongation of the snout may be associated with

a slight elongation of the mandibular bones and an

increase in the number of teeth they carry. The

diminutionof theeye with the small eye bulb freely

below the skin has not affected the skull as in

Astyanax fasciatus (Breder, 1944; cited by Culver,

1982).

4.4 Distribution, threats, and legal protection

Up to the present, Jelsevnik and DobliCica are the

only localities where P. a. parkelj has been found

with certainty. No reliable data exist on the finding

of Proteus in the mainspring of Jelsevnik. A report

on the finding of a black Proteusnear Mala Lahinja

(A. Hudoklin, pers. comm.), located 7 km to the

southeast of Doblicica and within the region of low

karst, has not been documented. No Proteus have

been reported from springs south of Doblicica (F.

Velkovrh, pers. comm.) or in any of the few active

caves south of Crnomelj. Unfortunately, all over

the Poljanska Gora the speleobiological sampling

conditions are unfavourable.

Black Proteus in populations of whites? The only

mention of the natural occurrence of "black"

specimens by Graf (1882) unfortunately is not ac-

companied by any morphological data. The par-

ticular karst spring in Slovenska Vas (formerly

Windischdorf) near Kocevje (Gottschee) in south-

eastern Slovenia has now been adopted for the

town's water supply. In recent years only white

specimens have been obtained from here.

White Proteus in populations of blacks? Local

inhabitants reported on the finding of white Pro-

teus in a locality between Jelsevnik and Doblicica.

Unfortunately no voucher specimens have become

available and the reports will have to be confirmed.

A voucher specimen does exist for a locality at

12 km to the northeast of Jelsevnik (Klepec, 1981;

cf. Fig. 1).

If the hypogean waters of the high karst of the

Poljanska Gora are its defined habitat, the exis-

tence of P. a. parkelj may be limited to water chan-

nels in the hydrographical background of both

springs that cover an area of approximately 55 km2

(cf. Habic et al., 1990). As has been suggested by


P. Habic (pers. comm.), the occurrence of P. a.

parkelj may in fact be restricted to the shallow karst

of the Bela Krajina while the deep karst of the

surrounding mountains is inhabited by the white

taxon. The available data do not contradict this

hypothesis. Indeed, the hypothesis is in line with

(unconfirmed) data about the sporadic occurrence

of white Proteus in Jelsevnik and black Proteus in

the springs of Lahinja. Because Proteus are dif-

ficult to find, their perceived absence in any area

does not provide a powerful argument.

The water from the Na Trati boiling holes at

Jelsevnik is polluted. The contaminationcan be at-

tributed to the dump of a smelting plant in a karst

doline uphill the Poljanska Gora, less than a

kilometre away. The fine grained and skin damag-

ing sandy debris dumped in the doline is rich in

phenols. It has been demonstratedthat (white) Pro-

teus in captivity did not survive a ten day period

when kept in close contact with these sands (T.

Valentincic, pers. comm.). We consider the pollu-

tion a serious threat to the survival of P. a. parkelj.

The boiling holes of Jamnice are hydrologically

similar but pollution has not been noticed there.

Since its habitat is not directly accessible to man,

a population estimate cannot easily be performed.

As it appears to be a rare taxon with a small distri-

bution area, P. a. parkelj has been placed on the

Slovenian Red List of endangered animal species

(Sket, 1992). It is under strict protection of the

Slovenianlaw as amemberof the category "perma-

nently cavernicolous animals", as well as in its own

right as a subspecies of P. anguinus. We discourage

collecting for any purpose. The most recent "Or-

dinance on protection of rare or endangered animal

species and their developmental stages" (Brelih &

Gregori, 1980) has become operational in 1976 and

a new act updating the old one is in preparation.

5 Acknowledgements

P. Habic, A. Mihevc, M. Zlokolica (Postojna), F. Velkovrh

(Ljubljana), and A. Hudoklin (Novo Mesto) provided valuable

information while M. Aljancic (Kranj), J.-P. Durand (Paris),

and H. Hagn (München) drew our attention to some literature

that otherwise would have gone unnoticed. F. Tiedeman (Vien-

na) provided specimens for biometrical analysis and Mrs. N.

Kranjec (Koper) performed biometrical studies that we used for

comparison. Mrs. V. Tabor and J. Subelj (Ljubljana) helpedus

out with the X-ray photography and A. Brancelj (Ljubljana)

identified Copepoda.A. Hodalic (Lausanne) made colour slides

that we reproduce and R. Griffiths (Canterbury) and A. Zuider-

wijk (Amsterdam) provided some of the outgroup material.

M. Garcia-Paris (Madrid) critically read the manuscript. Special

thanks are due to Silvo Medic and his family, inhabitants of the

hamlet of Jelsevnik, for promptly informingus about the local

hydrological conditions and for conscientiously collecting ani-

mals. Part of this study was funded by the Slovenian Ministry

for Science and Technology.

References

Aljancic, M., P. Habic & A. Mihevc, 1986. Crni moceril iz Bele

Krajine [The black Proteus from the White Carniola], Nase

Jame, 28: 39-44.

Aljancic, M. & B. Sket, 1964. Primer akcidentaine superregener-

acije pri mocerilu (Proteus anguinus Laur.) [Cas de super-

régéneration accidentelle chez le Protée Proteus anguinus

Laur.] Biol. Vest., 12: 109-113.

Arntzen, J.W.&G.P. Wallis, in press. The 'Wolterstorff Index'

and its value to the taxonomy of the Crested Newt superspe-

cies. Abh. Ber. Mus. Naturk. Vorgesch. Magdeburg, Special

Issue Wolterstorff Memorial Symposium.

Ayala, F. J., J.R. Powell, M.L. Tracey, C.A. Mouräo & Pérez-

Salas, 1972. Enzyme variability in the Drosophila willistoni

group, IV. Genie variation in natural populations of Dro-

sophila willistoni. Genetics, 70: 113-139.

Behler, J.L. & F. Wayne King, 1979. The Audubon Society field

guide to North American reptiles and amphibians: 1-743

(Alfred A. Knopf, New York).

Bowen, D.Q., G.M. Richmond, D.S. Fullerton, V. Sibrava,

R.J. Fulton & A.A. Velichko, 1986. Correlation of Quater-

nary glaciations in the northern hemisphere. In: V. Sibrava,

D.Q. Bowen & G.M. Richmond (eds.), Quaternary glacia-

tions in the northern hemisphere:509-510 (Pergamon Press,

Oxford).

Brelih, S. & J. Gregori, 1980. Redke in ogrozene zivalske vrste

v Sloveniji [Rare and endangeredanimal species in Slovenia]:

1—263 (Prir. Muz. Slov., Ljubljana).

Briegleb, W., 1962. Zur Biologie und Oekologie des Grotten-

olms (Proteus anguinus Laur. 1768). Z. Morph. Ökol. Tiere,

51: 271-334.

Bulog,B., 1991. Preliminarystudy of the sensory organs of the

Proteus sp. - black specimen (Urodela, Amphibia).The eye.

Abstracts Ordinary General MeetingSocietas Europaea Her-

petologica, 6: 20.

Bulog,B., 1992. Ultrastructural analysis of the retina of Proteus

sp.- dark pigmented specimens (Urodela, Amphibia). Elec-

tron Microscopy EUREM '92, 3: 659-660.

Conant, R., 1975. A field guide to reptiles and amphibians of

eastern and central North America (2nd ed.): i-xviii, 1 -429

(Houghton Mifflin, Boston).


Cooper, J.E. &R.A. Kühne, 1974. Speoplatyrhinuspoulsoni, a

new genus and species of subterranean fish from Alabama.

Copeia, 1974: 186-193.

Culver, D.C., 1982. Cave life: evolution and ecology: i-viii,

1-189 (Harvard Univ. Press, Cambridge Mass. & London).

Dolivo-Dobrovolsky, V., 1926. Lobanja cloveske ribice (Pro-

teus anguinus Laurenti)[Der Schädel des Grottenolmes (Pro-

teus anguinusLaurenti)].Rad JAZU, 232: 190-210,pis. 1-7.

Duellman, W.E.& L. Trueb, 1986. Biology of amphibians: i—-

xvii, 1-670 (McGraw-Hill, New York etc.).

Durand, J.-P., 1973. Développementet involution oculaire de

Proteus anguinus Laurenti urodèle cavernicole. Annls. Spé-

léol., 28: 193-208.

Durand, J.-P.& B. Delay, 1981. Influence of temperature on the

development of Proteus anguinus (Caudata: Proteidae) and

relation with its habitat in the subterranean world. J. thermal

Biol., 6: 53-57.

Estes, R. & I. Darevsky, 1977. Fossil amphibians from the Mio-

ceneofthe North Caucasus, U.S.S.R. J. paleontol.Soc'lndia,

20 ("1975"): 164-169.

Estes, R.D. & H.H. Schleich, in press. New material of Mio-

proteus caucasicus Estes and Darevsky from south German

localities (Amphibia: Caudata: Proteidae). Cour. Forsch. -

Inst. Senckenberg.

Farris, J.S., 1972. Estimating phylogenetic trees from distance

matrices. Am. Nat., 106: 645-668.

Fitzinger, L., 1850. Über den Proteus anguinus der Autoren.

Sitzungsber. Akad. Wiss. Wien math.-naturw. Cl., 2: 291 —

303.

Gams, I., 1965. Aperçu sur l'hydrologiedu Karst Slovene et sur

ses communications souterraines. Nase Jame, 7: 51-60.

Graf, E., 1882. Die Grottenwelt von Gottschee. Mitth. Sect.

Hoehlenk. oesterr. Touriste-Club, 1: 1-10.

Habic, P., 1991. Geomorphological classification of NW Di-

naric karst. Acta carsol., 20: 133-165.

Habic, P., J. Kogovsek, M. Bricelj & M. Zupan, 1990. Izviri

Doblicice in njihovo sirse krasko zaledje [Doblicica springs

and their wider karst background]. Acta carsol., 19: 5-100.

Harris, H. & D.A. Hopkinson, 1976. Handbook of enzyme elec-

trophoresis in human genetics (without consecutive pagina-

tion; North-Holland Publ. Co., Amsterdam).

Herre, W., 1935. Die Schwanzlurche der mitteleocaenen (ober-

lutetischen) Braunkohle des Geiseltales und die Phylogenie

der Urodelen unter Einschluss der fossilen Formen. Zoolo-

gica, 33 (87): 1-85.

Hodalic, A., 1993. Proteus: A la recherche du mystérieux

"poisson-humain". Animan, 42: 30-49.

Humason, G.L., 1967. Animal tissue techniques (2nd ed.): 1—

569 (W.H. Freeman and Co., San Francisco).

Istenic, L., 1987. O najdbierne cloveske ribice [On the finding

of a black proteus]. Proteus, 49: 243 -244.

Istenic, L. & B. Bulog, 1986. Crni moëeril se pod drobnogledom

[The black proteus under the microscope]. Delo, 25. Nov.

1986: 8.

Jenko, F., 1959. Hidrologija in vodno gospodarstvo krasa

[Hydrology and water management in karst]: 1-237 (Drzavna

zalozba Slovenije, Ljubljana).

Kammerer, P., 1912. Experimente über Fortpflanzung, Farbe,

Augen und Körperreduktion bei Proteus anguinus Laur.

Archiv Entwicklungsm., 33: 349-461, pis. I-IV.

Klepec, S., 1981. Cloveska ribica tudi v Beli krajini [Proteus also

in Bela Krajina], 3. dolenjski jamarski Tabor, Kostanjevica:

37-39.

Knauer, F., 1878. Naturgeschichte der Lurche (Amphibiologie).

Eine umfassendere Darlegung unserer Kenntnisse von dem

anatomischen Bau, der Entwicklung und systematischen Ein-

theilung der Amphibien sowie eine eingehende Schilderung

des Lebens dieser Thiere: i—xx, 1-340, 2 tables, 4 maps (A.

Pichler's Witwe & Sohn, Wien).

Kranjec, N., 1981. Razsirjenost in variabilnost mocerila (Pro-

teus anguinus Laurenti) [Distribution and variability of Pro-

teus anguinus]. Graduation thesis, Ljubljana: 1-70.

Lanyon, S.M., 1985. Detecting internal inconsistencies in dis-

tance data. Syst. Zool., 34: 397-403.

Maurer, H.R., 1971. Disc electrophoresis and related techniques

of Polyacrylamide gel electrophoresis (2nd ed.): i-xvi, 1-222

(De Gruyter, New York).

Maxson, L.R. & R.D. Maxson, 1979. Comparative albumin and

biochemical evolution in plethodontid salamanders. Evolu-

tion, 33: 1057-1062.

Melik, A., 1951. Pliocenska Pivka [The Pliocene Pivka River].

Geogr. Vestn., Ljubljana, 23: 17-39.

Melik, A., 1952. Zasnova Ljubljanicinegaporecja [Origins of

the Ljubljanicariver basin], Geogr. Zborn., Ljubljana, 1: 5—

31.

Melik, A., 1958. Jugoslavia: 1-675 (Tiskovna zadruga, Lju-

bljana).

Mertens, R. & L. Müller, 1940. Die Amphibien und Reptilien

Europas. Abh. senckenberg. naturf. Ges., 451: 1-56.

Nei, M., 1972. Genetic distance between populations.Am. Nat.,

106: 283-292.

Nei, M., 1987. Molecular evolutionary genetics: i-x, 1-512

(Columbia U.P., New York).

Novak, D., 1990. Novejsa sledenja kraskih voda v Sloveniji po

letu 1965 [Recent tracing of karstic waters in Slovenia]. Geo-

logija (Ljubljana), 33: 461—478.

Oosterbroek, P. & J.W. Arntzen, 1992. Area-cladograms of

circum-Mediterranean taxa in relation to Mediterranean

palaeogeography. J. Biogeogr., 19: 3-20.

Pehani, H. & A. Seliskar, 1941. O dozdevni metamorfozi

heteroplasticnihtransplantatovkoze neotenicnih amfibij [On

the supposed metamorphosis of heteroplastic skin grafts of

neotenic amphibians]. Zbornik prir. Drust., 2: 119-124.

Peters, N., G. Peters, J. Parzefall & H. Wilkens, 1973. Über

degenerative und konstruktive Merkmale bei einer phylo-

genetisch jungen Höhlenform von Poecilia sphenops (Pisces,

Poeciliidae). Int. Revue ges. Hydrobiol., 58: 417-436.

Poulik, M.D., 1957. Starch gel electrophoresis in a discontinu-

ous system of buffers. Nature, 180: 1477-1479.

Prelogovic, E., M. Arsovski, V. Kranjec, V. Radulovic, B.

Sikosek & S. Soklic, 1975. Paleogeografska evolucija teri-

torije Jugoslavije od tercijera do danas [Palaeogeographic


development of the Yugoslav territory since the Tertiary.]

Acta seismologica jugoslavica, 2-3: 7-11, maps.

Rafinski, J. &. J.W. Arntzen, 1987. Biochemical systematics of

the Old World newts, genus Triturus: allozyme data. Her-

petologica, 43: 446-457.

Rogers, J.S., 1972. Measures of genetic similarity and genetic

distance. Stud. Genet. VII, Univ. Texas Publ., 7213: 145—

154.

Romer, A.S. & T.S. Parsons, 1977. The vertebrate body (5th

ed.): i-viii, 1-624 (Saunders, Philadelphiaetc.).

Schreiber, E., 1912. Herpetologia europaea. Eine systematische

Bearbeitung der Amphibien und Reptilien welche bisher in

Europa aufgefunden sind (2nd. ed.): i-x, 1-960 (Gustav

Fischer, Jena).

Shaw, C.R. & R. Prasad, 1970. Starch gel electrophoresis ofen-

zymes - a compilation of recipes. Biochem. Genet., 4: 297—

320.

Sket, B., 1970. Über Struktur und Herkunft der unterirdischen

Fauna Jugoslawiens. Biol. Vestn., 18: 69-78.

Sket, B., 1983. Je Marifugia res ubeznik iz morja? [Is Marifugia

really a fugitive from the sea?] Proteus, 46: 102-104.

Sket, B., 1985. Why all cave animals do not look alike - a dis-

cussion onadaptive value of reduction processes. NSS Bull.,

National Speleological Society, 47: 78-85.

Sket, B., 1986. Evaluation of some taxonomically, zoogeo-

graphically, or ecologically interesting finds in the hypogean

waters of Yugoslavia (in the last decades). 9. Congr. int.

Espeleol., Comun., 2: 126—128.

Sket, B., 1992. Rdeci seznam ogrozenih vrst dvozivk (Amphibia)

v Sloveniji [The Red List of endangered Amphibia in Slove-

nia]. Varstvo narave, 17: 45-49.

Sket, B. & J. Bole, 1982. Organisms as indicators of hypogean

water connections (summary). Naä Krä. (Sarajevo), 6: 115—

117.

Sket, B. & F. Velkovrh, 1978. The discovery of Proteus-eggs

(Proteus anguinusLaurenti, Amphibia) in seminatural condi-

tions. Int. J. Speleol., 10: 205-209.

Sneath, P.H.A. & R.R. Sokal, 1973. Numerical taxonomy. The

principles and practice of numerical classification: i-xv,

1-573 (W.H. Freeman, San Francisco).

Swofford, D.L. & R.B. Selander, 1981. BIOSYS-1: A FOR-

TRAN program for the comprehensive analysis of electro-

phoretic data in population genetics and systematics. J.

Hered., 72: 281-283.

Valvasor, J.W., 1689. Die Ehre des Herzogthums Crain: 1-696

(W.M. Endtner, Nürnberg).

Vandel, A., 1965. Biospeleology. The biology of cavernicolous

animals: i-xxiv, 1-524 (Pergamon Press, Oxford).

Vandel,A. & M. Bouillon, 1959. Le Protée et son intérêt biolo-

gique. Annls. Spéléol., 14: 111-127.

Vandel, A., J. Durand & M. Bouillon, 1966. Contribution à

l'étude du développement de Proteus anguinus Laurenti

(Batraciens, Urodèles). Annls. Spéléol., 21: 609-519.

Wallis, G.P. & J.W. Arntzen, 1989. Mitochondrial-DNA varia-

tion in the crested newt superspecies: limited cytoplasmic gene

flow among species. Evolution, 43: 88-104.

Wilkinson, L., 1989. SYSTAT: The system for statistics: manu-

al for computer programme (SYSTAT Inc., Evanston, IL.).

Wolterstorff, W., 1923. Übersicht der Unterarten und Formen

des Triton cristatus Laur. Blätter Aquar. Terrarienk. Stutt-

gart, 34: 120-126.

Received: 29 June 1993

Revised: 17 November 1993

proteus anguinusparkelj (urodela: proteidae) project/a black...

Documents